To salvage (routinely) or not to salvage: that is the question

To salvage – it means to save – from a shipwreck, a fire, a wrecking ball or other forms of permanent destruction or loss. The act of salvaging implies repurposing, putting to use somewhere else. It makes perfect sense in a world that throws away and loses a lot of things. Cell salvage also makes perfect sense: collect the patient's own blood from the operative field, wash away the impurities and re-infuse it to the patient; thus, avoiding the risks, and costs, of allogeneic transfusion. But a lot of things in medicine that appear to make perfect sense do not actually work as expected. For example, the practice of bloodletting to rid the body of bad humours made perfect sense to our medical colleagues practicing in previous centuries, but it does not work and may cause harm. Similarly, we need to ask, does cell salvage use actually result in lower allogeneic transfusion rates and its associated risks? Specifically, in the obstetric population, does the benefit of routine use of cell salvage outweigh the risks? Postpartum haemorrhage is the leading cause of maternal mortality worldwide. The incidence is increasing, and the incidence of postpartum haemorrhage requiring blood transfusion in the US nearly quadrupled between 1993 and 2014 1. In 2014, 0.4% of all deliveries required a blood transfusion. Importantly, experts estimate that between 70% and 90% of postpartum haemorrhage-related deaths are preventable 2, 3. The morbidity of postpartum haemorrhage often reflects the quality of the clinical response; early recognition and prompt management of haemorrhage are critical to improving maternal outcomes. Recommendations to improve care include: identification of women at high risk for haemorrhage; improved recognition of haemorrhage; and timely management. Yet, identifying at-risk women is challenging; many women who experience postpartum haemorrhage do not have identifiable risk factors 4. Therefore, strategies such as the routine use of cell salvage during caesarean section might improve the timeliness of clinical response and improve outcomes. But does this strategy actually work, and is it safe and cost effective? In this issue of Anaesthesia, Sullivan and Ralph report the results of a retrospective observational study of one UK centre's experience with cell salvage in 6352 obstetric patients 5. Between 2008 and 2017, cell salvage was used routinely for 98% of caesarean deliveries. Salvaged blood was re-infused in 1170 women (18.4%), and only 44 women (3%) required an allogeneic blood transfusion in addition to salvaged blood. In a subset of patients, the investigators assessed the ‘quality’ of the salvaged blood; fetal blood cells were found in all samples; however, the rate of allo-immunisation was very low. Of note, the authors used a single suction catheter to aspirate blood from the surgical field; washed all surgical swabs; and, in the latter years of the study period, did not use a leucocyte depletion filter. They reported no major safety events. The authors concluded that “it is possible to use cell salvage at caesarean section both safely and economically” without the need for additional staff, using one suction device and no leucocyte depletion filter. Although these results are promising, they contrast with the results of the cell salvage in obstetrics (SALVO) trial, a pragmatic, multicentre (26 UK units), randomised controlled trial 6. In the trial, 3028 women at risk for haemorrhage and undergoing a scheduled or intrapartum caesarean section were randomly allocated to receive cell salvage in addition to standard care (experimental group) or standard care alone (control group). The rate of allogeneic transfusion was 2.5% in the cell salvage group and 3.5% in the standard care group, a difference that was not statistically different (adjusted risk difference −1.03, [95%CI: −2.13 to 0.06]) 6. On average, 1 in every 100 women who received routine cell salvage avoided an allogeneic transfusion. Similar to the study by Sullivan and Ralph 5, the SALVO trial investigators concluded that use of cell salvage in obstetric patients was safe 6. Finally, the SALVO investigators used their data to perform a cost effectiveness analysis from the perspective of the NHS, and concluded that cell salvage was more expensive than standard care; the incremental cost effectiveness ratio was £8110 (€9711; $10,303) per transfusion avoided 6. How can we bridge the gap between the Sullivan and Ralph study and the SALVO trial 5, 6? The studies differed methodologically. In the hierarchy of evidence, the findings of randomised controlled trials are normally considered superior to observational studies. The SALVO trial was a large, well-conducted, randomised controlled trial in women at increased risk for haemorrhage 6. The primary outcome was the rate of women receiving allogeneic blood transfusion to manage haemorrhage; secondary outcomes included safety and cost. In contrast, the Sullivan and Ralph study was a retrospective observational trial 5. All women undergoing caesarean section were included. Although outcomes were not defined a priori, the authors reported allogeneic transfusion rates over the 10-year study period, as well as safety data. Although Sullivan and Ralph did not report cost as an outcome of their study, they emphasised key differences between their own practice and that described in the SALVO study 5, 6. Specifically, in the SALVO study, a full setup for both collection and processing of salvaged blood was mandated as part of the study protocol 6. In contrast, in the Sullivan and Ralph study, the cell salvage machine was initially setup for collection only, and the processing kit was opened only if a sufficient amount of blood had collected in the reservoir 5. By protocol, all surgical swabs were washed and a single suction device was used to aspirate blood from the surgical field, rather than using one suction before delivery of the placenta, and a second one afterwards (the use of two suction catheters has been recommended to reduce the volume of amniotic fluid collected into the cell salvage reservoir 7). Leucocyte depletion filters were not used after 2015, and perhaps most importantly, no additional staff were required to operate the cell salvage machine. Operating department practitioners, who are routinely available to assist the anaesthetist, were trained in cell salvage operation and were available 24 h a day, 7 days a week. In contrast, in the SALVO trial, many centres required additional personnel to operate the cell salvage machine, and just over half of the centres used a leucocyte depletion filter, washed swabs and used one suction, rather than two 6. Thus, personnel costs and routine use of the processing kit were likely to have been significant costs in the SALVO trial. Importantly, the cost-saving measures used by Sullivan and Ralph did not appear to reduce the quality of the salvaged blood, or result in any identifiable patient harm, although both studies were underpowered for safety outcomes 5, 6. Although it was once thought that a leucocyte depletion filter was necessary to reduce amniotic fluid contaminants in the salvaged blood 7, many centres, including nearly half of the centres in the SALVO trial 6, have abandoned their use due to reports of acute hypotensive events, as well as slower infusion rates 8. Indeed, 16 out of 18 adverse events reported in the SALVO trial were thought to be related to the leucocyte depletion filter 6. An additional safety concern is maternal allo-immunisation. Among the subset of 647 women in the Sullivan and Ralph study who had subsequent allo-immunisation testing, only two women developed new antibodies 5. Additionally, the authors reported that, in the past 3 years, they observed no increase in the dose of Rho(D) immune globulin administered to Rh-negative women with Rh-positive fetuses, indirect evidence that abandoning the use of leucocyte depletion filters does not increase the risk of fetomaternal haemorrhage 5. In contrast, the SALVO investigators reported a greater incidence of fetomaternal haemorrhage, defined as ≥ 2 ml, in women who received cell salvage compared with those who did not 6. Although the Sullivan and Ralph study was not powered to investigate this safety outcome, the results add reassurance that leucocyte depletion filters may not be necessary for routine cell salvage 5. More studies are needed to fully understand the impact of cell salvage reinfusion on the risk of allo-immunisation. When making the decision to use any medical intervention, the benefits must be weighed against the risks, including cost. Arguably, the primary benefit of cell salvage is a reduction in allogeneic blood transfusion. The SALVO study found no benefit in terms of reduced allogeneic blood transfusion when cell salvage was used routinely for high-risk patients undergoing caesarean section, although in a planned sub-group analysis in women undergoing emergency caesarean section, the allogenic transfusion rate was lower in the cell salvage group at 3.0% vs. 4.6% in the standard care group 6. Although Sullivan and Ralph reported an inverse relationship between cell salvage use and allogenic transfusion over the 10-year study period, the causality of this association must be viewed with scepticism 5. As noted by the authors, other blood conservation strategies were also implemented during the study period that were likely to have contributed to the observed decrease in the allogeneic transfusion rate. However, before we totally dismiss the routine use of cell salvage for obstetric patients, we should consider other possible benefits of its use. These include improved ability to rescue, a reduction in maternal near-miss events and death due to haemorrhage, a leading, but usually preventable cause of maternal mortality. Perhaps, the rate of allogeneic blood transfusion is not the outcome we should be assessing, but rather other important outcomes, such as failure to rescue. As confirmed in the current study as well as in others, the risks of cell salvage are low and the costs could be lower if modelled on the setup described by the authors in their institution 5. The SALVO study cost analysis only considered direct costs from the provider perspective – it did not consider the costs, to both the provider and to society, of serious adverse outcomes such as failure to rescue. The death of a mother due to haemorrhage is a tragedy, all the more so because it is almost always preventable with appropriate and timely management. In the US, the National Partnership for Maternal Safety, an organisation formed with the goal of reducing maternal mortality, published an obstetric haemorrhage safety bundle in 2015 9. A key element of the bundle is readiness. It is possible that using cell salvage in all patients improves the team's preparedness to manage cell salvage, the quality of the salvaged blood and perhaps other components of preventing and treating postpartum haemorrhage. The routine use of cell salvage would certainly improve our ability to initiate treatment early in haemorrhage. It would provide the ‘regular experience’ required by the National Institute for Health and Care Excellence (NICE) guidelines for its use 10. Furthermore, its use might serve as a trigger to focus the team on postpartum haemorrhage diagnosis and treatment, leading to improved outcomes. Many organisations, including the Royal College of Obstetricians and Gynaecologists, the National Institute of Health and Care Excellence (NICE), the American College of Obstetricians and Gynecologists and the American Society of Anesthesiologists, recommend that cell salvage be considered in selected situations, such as when large blood loss is expected (> 20% of blood volume), when patients refuse allogeneic blood or during intractable haemorrhage when banked blood is not available 10-13. These guidelines imply, although they do not directly state, that the routine use of cell salvage in all obstetric patients is not indicated. Indeed, in the first updated guideline published since the SALVO trial, the Association of Anaesthetists state that cell salvage “is not used routinely for caesarean section based on the current evidence, be it elective, urgent or emergency” 14. We suggest that further research is necessary before completely condemning the routine use of cell salvage. The sub-group analysis in emergency caesarean section patients in the SALVO study suggests that the technique may have benefit in this patient group 6. We suggest a randomised controlled trial of the routine use of cell salvage in emergency caesarean section patients, incorporating the paradigm suggested by Sullivan and Ralph, and powered to assess outcomes such as near misses or a composite outcome of adverse outcomes, is a rational next step for further defining the role of cell salvage in obstetric care. No external funding or competing interests declared.